---
type: "Journal Guide"
title: "Why the Same Carnatic Song Can Sound Different: A Listener’s Guide to Pāṭhāntara and Present-Day Versions"
description: "Learn how to compare Carnatic song versions through sahitya, melodic phrase, tempo, lineage and source, while keeping evidence limits visible."
resource: "https://vidyabhushana.com/blog/carnatic-pathanthara-listening-guide.html"
language: "en-IN"
author: "Living Sabha Editorial Desk"
date_published: "2026-09-19"
date_modified: "2026-09-19"
tags:
  - "pāṭhāntara"
  - "pathanthara"
  - "Carnatic music"
  - "sahitya"
  - "oral versions"
  - "sampradaya"
  - "song versions"
  - "listening guide"
  - "documentation"
generated: { by: "process:living-sabha-static-build", at: "2026-09-19T00:00:00Z" }
verified:
  - { by: "human:living-sabha-editorial-desk", at: "2026-09-19T00:00:00Z" }
status: stable
sources:
  - id: "reference-1"
    resource: "https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf"
    title: "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract"
  - id: "reference-2"
    resource: "https://musicacademymadras.in/academic-sessions-16th-december-2025-to-1st-january-2026/"
    title: "The Music Academy, Madras, Academic Sessions, 16 December 2025 to 1 January 2026"
  - id: "reference-3"
    resource: "https://www.rasikas.org/forums/viewtopic.php?p=386126"
    title: "Rasikas.org, Annamayya krithi song swaras, listener request posted 5 September 2026"
  - id: "reference-4"
    resource: "https://www.rasikas.org/forums/viewtopic.php?p=386129"
    title: "Rasikas.org, Annamayya krithi song swaras, reply posted 6 September 2026"
  - id: "reference-5"
    resource: "https://www.rasikas.org/forums/viewtopic.php?p=386138"
    title: "Rasikas.org, Annamayya krithi song swaras, follow-up posted 7 September 2026"
---

# Why the Same Carnatic Song Can Sound Different: A Listener’s Guide to Pāṭhāntara and Present-Day Versions

> This Markdown file is an alternate reading copy of the canonical HTML guide. Use the canonical HTML page for citation, presentation, and the latest public rendering.

A careful guide to comparing documented present-day versions of a Carnatic composition through text, melodic phrase, tempo, lineage statement, and source, without declaring one version correct or incorrect.

## A short answer

A Carnatic composition can sound different from one present-day rendering to another because a title does not carry every detail that reaches the ear. Listeners may encounter a different wording, a different division of words, a different melodic turn, a different speed, a stated connection to a different teacher or school, or simply a different kind of source. None of those observations automatically establishes that one rendering is right and the other is wrong. The useful first task is to describe what is actually available, then to say what the available material does not establish.

This guide uses **pāṭhāntara** as a practical comparison word. It helps a listener speak about documented oral or present-day versions without turning a comparison into an authenticity contest. The comparison has five separate fields: **song text**, **melodic phrase**, **tempo**, **lineage statement**, and **source record**. Keeping those fields apart prevents a common error. A listener may hear a changed phrase and then assume that the text, raga, lineage, and history must also have changed. The recording may not support any of those further conclusions.

The Music Academy abstract for Rama Ravi’s 24 December 2025 session offers a helpful, bounded frame. It proposes questions about differences in sahitya, melody, and speed among oral versions of selected Muttusvāmi Dīkṣitar compositions. It also asks how sampradaya and raga treatment relate to such differences. The abstract presents those questions for a defined repertory and intends audio demonstration. It does not announce a universal rule for every Carnatic song. [ref-1](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract") This guide follows the discipline of the questions, not an answer key.

The result can be modest but useful. A listener may be able to say that two recordings use different audible phrase cues, move at different overall paces, and carry different source notes. The listener may also have to say that neither recording proves an earlier text, an original tune, a composer’s intention, or a transmission history. That is not a failure of listening. It is an honest description of the evidence in hand.

## Start with comparison, not a verdict

When two performances carry the same title, the temptation is to make a quick judgment. One may sound familiar and the other unfamiliar. One may match a class note, a printed lyric, a family memory, or a favourite recording. Familiarity can be emotionally important, but it is not itself a source category. It tells the listener that a version is known to them. It does not tell the listener how the version was transmitted, which words have documentary support, or whether another rendering is invalid.

A comparison begins more calmly. Give the two items neutral labels such as **Version A** and **Version B**. Record the date, platform, performer or uploader if credited, and the exact place where each item was found. Write only what can be heard or read at this stage. If Version A repeats a word where Version B moves on, that can be noted. If one opening phrase seems longer, that can be noted. If a visible caption supplies a raga or a teacher’s name, record that as a caption statement rather than silently treating it as established history.

This restraint matters because several kinds of question can hide inside the word “same.” The title may be the same while the text is not identical. The text may look identical while the musical setting differs. The melodic setting may be close while the speed differs. A performer may explicitly name a lineage connection, while another recording may provide no lineage information at all. These are different observations. A single label such as “changed” cannot do their work.

The aim is not to make listening bloodless. Careful description can sharpen attention. It lets a listener return to a particular word, phrase, or cadence and notice it more precisely. It also leaves room for qualified language: “audibly different in this excerpt,” “stated in the upload,” “not supplied by the source,” or “not established by the two recordings.” Such phrases are signs of accuracy, not evasions.

## What pāṭhāntara means in this guide

In this guide, **pāṭhāntara** is not used as a stamp that certifies a version. It is a way to name the problem of more than one oral or present-day rendering of a composition. The Music Academy abstract describes pāṭhāntara as oral versions in the Dīkṣitar sampradaya and asks how similarities and differences may be examined in selected compositions. [ref-1](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract") That limited institutional usage supports a listener’s basic starting point: versions can be compared, and the comparison may involve more than one musical feature.

The word should not be made to carry claims it cannot support. A listener who hears two versions does not thereby identify a complete sampradaya. A performer’s stated teacher connection is not the same thing as a verified transmission map. A notation book is not the same thing as a recording. A recording from a particular date is not the same thing as a historical witness to the first form of a composition. These distinctions do not diminish oral practice. They make the listener’s record clearer about what kind of evidence is being used.

It is also safer to avoid using pāṭhāntara as a polite substitute for “correct version.” That move merely hides the verdict. A version can be carefully described without an approving or disapproving adjective. For example, “Version A holds the final syllable longer in this passage” is a documented listening note. “Version A preserves the authentic text” needs much more evidence than an audible hold. The difference lies in the kind of statement, not in the listener’s confidence.

For a new listener, the practical benefit is simple. Pāṭhāntara invites close hearing while protecting against overreach. It asks, “What is different here?” before it asks, “Why is it different?” It permits the answer “the reason is not supplied.” It also makes space for a version to have its own present-day coherence without turning that coherence into proof about the past.

### A song title is not a complete version record

A title is a useful doorway, but it is a thin description. It can help a listener locate a composition, a lyric page, a concert list, a recording, or a discussion. It cannot by itself identify the exact text, the tune, the phrase shapes, the tempo, the raga label, the tala label, the teacher connection, or the date of a particular performance. Treating a title as if it held all of those facts creates avoidable confusion.

Build the habit of asking, “Which instance of this title am I hearing?” An instance might be a studio recording, a concert upload, a classroom clip, a printed edition, a handwritten note, or a social post with a caption. Each instance has a different amount of visible context. A complete comparison begins by naming the instance, rather than beginning with an abstract song title. This makes later corrections possible. Another listener can return to the same material and test the observation.

The distinction is especially important when a listener is working from an excerpt. An excerpt may omit the beginning, remove the surrounding composition, cut a repeated phrase, or alter the perceived pace. It may be compressed or edited. The listener can still record an audible cue, but should describe it as a cue in an excerpt. They should not expand it into a complete account of the composition. A small source can support a small claim.

A version record therefore needs a date and a location, not only a title. It needs a performer or uploader name when supplied, not an inferred identity. It should preserve the original spelling found in the source and also note variant spellings when they affect retrieval. These small administrative steps may seem distant from music, yet they are what allow musical observations to remain checkable. Documentation does not replace listening. It gives listening a stable address.

## The five fields that keep a comparison clear

The most practical way to compare two present-day versions is to open five columns before pressing play. The first column is **song text**, often called sahitya when discussing the words. The second is the **melodic phrase**. The third is **tempo or pace**. The fourth is the **lineage statement**, meaning a connection to a teacher, school, or sampradaya only when a source actually states one. The fifth is the **source record**, which identifies what the item is and where its information appears.

These fields are deliberately separate. The text field can contain a visible spelling or an audible word division. The phrase field can contain a descriptive contour, a time range, or a note that a phrase was not clear enough to describe. The tempo field can record a relative observation such as “the refrain returns more quickly in this clip,” without pretending to give a precise measurement. The lineage field can contain an exact credited statement or “not supplied.” The source field can contain a URL, edition details, a programme entry, or a note that the item is an uncredited repost.

A five-field record slows down a familiar mistake. A listener sometimes sees a raga label in one source and then writes it under the phrase field. Another listener hears a fast delivery and then calls it a lineage feature. A third sees a famous performer’s name and infers a complete textual authority. Each move collapses unlike evidence. The fields do not remove uncertainty. They stop uncertainty from being disguised as a single confident story.

The table below is not a ranking device. It is a prompt for clear observation. Neither side has a score. The final row is intentionally open. A comparison is complete even when important questions remain unanswered.

Field

Version A: record only what is available

Version B: record only what is available

Safe conclusion

Song text

Audible or printed word cue, with location

Audible or printed word cue, with location

Text cues are same, different, or unclear in the compared material

Melodic phrase

Time range and plain contour note

Time range and plain contour note

Phrases sound alike, different, or cannot be compared from the source

Tempo

Relative pace in the same passage

Relative pace in the same passage

Pace differs in these instances, not necessarily in the composition as a whole

Lineage statement

Exact stated credit, or “not supplied”

Exact stated credit, or “not supplied”

A source states a connection, or does not state one

Source record

Recording, edition, caption, programme, or note

Recording, edition, caption, programme, or note

The two witnesses have different kinds and amounts of context

Unknown

Questions not answered by the item

Questions not answered by the item

Unresolved questions remain visible

/manus-storage/pathanthara-comparison-card\_1f9902fd.png

## Field one: Song text and sahitya

Text is often the first difference a listener notices. A word may be pronounced with a different division. A repetition may appear in one performance and not another. A syllable may receive more musical space. A performer may use a word that does not match the listener’s lyric sheet. Before calling any of these a textual change, note what kind of access you have. Are you hearing a vocal recording? Reading a printed source? Looking at a caption? Working from an automatic transcript? Each route has its own limits.

A careful text note preserves the evidence. Write the exact word or short phrase as it appears in the visible source, or mark it as a tentative hearing. Include a time range for an audio note. If the language is not known to the listener, that should be visible too. A listener need not pretend to resolve every syllable. “Unable to distinguish the word in this audio” is more useful than a confident but uncheckable spelling.

Do not treat musical emphasis as automatic evidence of a different text. A held syllable may change the listener’s experience without changing the underlying word. Conversely, two printed texts may look identical while the sung division creates a different contour. The text field should record words and divisions. The phrase field should record the musical treatment. Keeping them apart allows later comparison without forcing a false choice between literature and sound.

The guide also avoids silently correcting a version against a favourite source. A source can be described as an edition, a lyric page, a class handout, or an upload. It should not be called definitive unless that authority is directly demonstrated within the material under review. If text sources conflict, write the conflict down. State which source supplies each reading. Do not use the listener’s preference to bridge a documentary gap.

A useful text record has two layers. The first layer is **what the source makes available**. This might be a printed line, a caption, a subtitle, a transcription supplied by the uploader, or words heard in a vocal passage. The second layer is **how confidently the listener can report it**. A printed edition may show a spelling clearly but leave its editorial basis unknown. A recording may make a word sound clear to one listener and ambiguous to another because of diction, accompaniment, room sound, or the listener’s language knowledge. A good note does not make those layers disappear. It writes them down separately.

Suppose that one version appears to repeat a word and another seems not to. Start with the narrowest description: “In the marked passage, Version A contains a second audible occurrence of this word; Version B does not contain a clearly comparable occurrence in this excerpt.” This sentence does not assume that Version B deletes the word from the composition. It does not assume that Version A is using a fuller text. It identifies an event in a marked passage. The wording can later be revised if a complete recording or a text source changes the picture.

Word division deserves equal care. A sequence of syllables can be distributed differently across the musical line. What first sounds like a changed word may be a changed join between words. What first sounds like a missing word may be covered by an instrumental response or a breath. Listeners who do not know the language should resist compensating with a familiar printed lyric. The printed lyric may help formulate a question, but it should not be silently substituted for what the recording audibly contains. A clear sheet can hold both: “printed text reads this” and “audio at this point is unclear.”

Transliteration is another field where a record can become misleading if it hides its source. If an upload provides Roman letters, record that it does. If the listener produces a practical transliteration to keep track of a passage, label it as a listening aid rather than as an edition. The same care applies to translation. A translation can clarify a listener’s own note, yet it may introduce interpretive choices not audible in the recording. The present guide does not need a translation to compare whether a repeated sound cue occurs. When meaning matters, the source and the translator should be identified.

Text comparison also benefits from recording **location** rather than building a free-standing list of variants. A word has a role in a sequence. It may occur before a refrain, after an instrumental response, or at the close of a line. A location note helps another listener align the material. It is more reliable than a detached statement that “the lyric is different.” If the recordings are long, make a short clip note for each observation. If the recordings are short, say that the context is short. The scale of the claim should match the scale of the access.

When printed sources differ, do not create a synthetic text by selecting a convenient word from each. That produces a third text that neither source presents. Instead, identify the witnesses. A neutral note can say, “Source X prints this form; Source Y prints that form; the relationship between them is not explained in the materials consulted.” This procedure may feel slow. It is faster than correcting a public claim after it has absorbed an unsupported mixture of sources.

The listener should also distinguish a **textual observation** from an **attribution claim**. A page may place a composer’s name beside a lyric. That page is evidence that it displays the name. It is not, without further evidence, a complete attribution argument. In the same way, the listener should not identify a language or dialect from an uncertain audio excerpt merely because a song is commonly associated with a language. A title, a label, a word, and an attribution each deserve their own source note.

This carefulness does not require a listener to become silent. It gives the listener a way to speak accurately. “The audio seems to use this word, but I need a reliable text source to confirm it” is an excellent public note. “The source writes this phrase, while the vocal articulation at this time range is not distinct enough for me to compare every syllable” is equally helpful. Both sentences preserve the difference between hearing, reading, and knowing.

Finally, the text field can make later musical comparison more humane. Listeners sometimes hear an unfamiliar word and treat it as a fault. A documented record replaces that reflex with a question: what source carries this word, and what does the available material let us conclude? The answer may be incomplete. Even so, the note has moved the conversation from correction to inquiry. That is the right foundation for comparing present-day versions.

## Field two: The melodic phrase

A melodic phrase is the part of the comparison that listeners may recognize most quickly and describe least carefully. It is easy to say that one version has “different swaras” or “a different raga” after hearing a changed turn. Such conclusions go beyond what a short excerpt can usually bear. A safer first note names an audible event: a rise arrives later, a return seems shortened, a held point is absent, or an ornamental movement is more pronounced in one recorded passage.

The phrase field does not ask the listener to publish a full swara transcription. It asks for a traceable cue. A time range is often enough. So is a simple phrase description in plain language. “The phrase after the repeated word descends before the cadence in Version A; the comparable moment in Version B moves directly to the cadence” is a useful note when it is tied to the material. It remains useful even if the listener cannot name every pitch.

This approach is compatible with the Journal’s guide to [how to read Carnatic music notation](https://vidyabhushana.com/blog/how-to-read-carnatic-music-notation.html), but it is not the same task. Reading a credited notation is different from deriving authoritative notation from a recording. The latter may demand contextual knowledge not contained in the clip. A comparison sheet should therefore say whether it relies on a printed phrase, a credited notation, or an audible description. Those are not interchangeable forms of evidence.

The same caution applies to gamaka. The guide on [what gamaka is in Carnatic music](https://vidyabhushana.com/blog/what-is-gamaka-in-carnatic-music.html) can help a listener notice that a note is approached or shaped in motion. It does not allow the listener to infer a complete lineage or raga history from one ornament. In a comparison sheet, describe the audible shape first. If a trusted source names it, record the source next to that name. If no such source is supplied, leave the specialist label out.

## Field three: Tempo and pace

Tempo is a real listening feature, but it needs a humble vocabulary. One version may feel more spacious. Another may move more quickly through a repeated line. An accompanimental texture may make one rendition seem urgent even when the underlying pulse is not easily comparable. A listener can hear and record these differences. The listener should not assume that speed alone identifies an earlier or later form, a correct form, or a particular school.

The Music Academy abstract explicitly includes a question about noticeable speed variations in the renditions it proposes to examine. [ref-1](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract") That is a useful reminder that pace deserves a field of its own. It is also a warning against folding pace into melody. A slower rendering may make a phrase sound more elaborate to the ear, while a faster rendering may make the same broad contour harder to isolate. If the listener calls the phrase different without accounting for pace, the comparison may be distorted.

Use a matched passage whenever possible. Compare the same refrain, the same lyric cue, or another identifiable location. Avoid comparing a free opening from one item with a metrical refrain from another and then calling the contrast a tempo difference. If the sources do not provide comparable passages, write “not comparable from the available extracts.” Precision begins with matching the question to the evidence.

A listener can add time stamps without turning the worksheet into a laboratory report. “Version A takes roughly longer to return to the repeated line in this excerpt” is a description. If a numerical duration is noted, it should refer only to the marked segment. It should not be converted into a claim about the performer’s general style or the composition’s intended speed. The source may show an instance, not a norm.

## Field four: Lineage statements and sampradaya

Lineage is often discussed with warmth and confidence, yet it is among the easiest fields to overstate. A recording may name a teacher. A programme note may name a school. A performer may speak about a tradition. Another recording may offer no such information. The responsible worksheet records these as **statements in a source**. It does not transform a visible name into a complete and verified chain of transmission.

The Music Academy abstract uses sampradaya while asking whether differences are exclusive to the Dīkṣitar sampradaya or are also found in others. [ref-1](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract") This is a nuanced question, not a shortcut. It does not say that any audible difference can be assigned to a sampradaya. It does not say that every performance with a teacher credit represents a distinct pāṭhāntara. Its value for a listener is methodological. The source of a connection must be marked, and the relationship between connection and sound should not be assumed.

Write the statement as exactly as the available material allows. “Programme credits the artist as a student of X” is different from “this phrase comes from X’s lineage.” The first can be checked against the programme. The second would require a source that actually makes that musical claim. If the wording is ambiguous, preserve the ambiguity. A note such as “teacher name appears, but the source does not link it to this version” keeps the record honest.

Lineage should not become a prestige column. It is not a scorecard. It can explain where to look next, perhaps toward a credited teacher, edition, archive, or programme note. It may also remain blank. An empty field does not mean a version has no transmission. It means that the present source does not document that transmission. This distinction respects both the music and the limits of public records.

## Field five: The source record

A source record answers a simple question: what exactly is being compared? It should name the form of the witness before it makes a musical claim. A concert recording, a studio release, a social-media clip, a printed score, a lyric page, a programme listing, a forum post, and a listener’s memory do not carry the same evidentiary weight for the same question. They can all be useful. They must not be made equivalent.

For every item, record the source type, the visible date if supplied, the title as presented, the credited performer or uploader if supplied, and a stable link or bibliographic detail. If a source has no date, say so. If it has been reposted by an account that does not name the original performer, say so. If an excerpt has no beginning or end, say so. These notes help prevent a later reader from confusing a circulating clip with a complete, credited performance.

The source field is where a listener separates a claim from its carrier. “The page labels the item with a raga” belongs in the source record as a page label. “The item sounds like that raga” is a different claim, and this guide does not ask a listener to settle it from a label alone. Likewise, “the recording includes a phrase at 01:14” is a recording observation. “The phrase was always sung this way” is historical language that the recording cannot supply.

Good source records make disagreement more productive. Two listeners may hear the same passage differently, but they can at least identify the same audio and time range. A researcher may later bring a better edition, a programme note, or a documented archive to the question. The original note remains useful because it states its basis. It has not been inflated into a conclusion that blocks later evidence.

## What a recording can document

A recording is a strong witness to the event it preserves. It can document that an audible passage occurs in that recording. It can support a time-stamped description of a repeated word, a phrase cue, a tempo impression in a matched excerpt, an instrumental entry, or a visible performance credit. A recording can also show what contextual information its own caption, sleeve note, or uploader supplies. These are valuable forms of evidence when their scope is stated clearly.

The phrase “in this recording” does important work. It distinguishes an event from a general rule. A listener may write, “In this recording, the repeated line is heard after the instrumental response.” That statement can be checked by returning to the audio. It does not claim that all performances place the line there, that the composition requires that order, or that the arrangement is inherited from a particular teacher. The claim stays as large as the evidence.

A recording may also be compared with another recording on equal terms. The listener can align similar textual cues, mark time ranges, and describe audible differences. Such a comparison is often more illuminating than a vague claim that one version “feels traditional.” It produces a small record that another listener can revisit. It leaves room for the sound to be rich without requiring the note to be grand.

Audio quality affects the strength of an observation. Compression, crowd sound, editing, microphone placement, and partial clips can hide words and blur phrase edges. A responsible note can include “not clear enough to distinguish” or “excerpt begins after the relevant lead-in.” These are not apologies. They tell the next listener how much confidence the record deserves. The aim is not exhaustive certainty. It is a transparent account of a particular audible witness.

## What a recording cannot settle

A recording does not by itself establish an original form, a composer’s intention, the authority of a text, or a complete transmission history. It may be old or recent, carefully produced or casually uploaded, famous or obscure. Its status changes what questions it can help answer, but an audio file alone remains evidence of a performance or recording event. It cannot speak beyond the trace it preserves without additional documentation.

This limit is especially important when listeners use terms such as “original,” “traditional,” “pure,” or “correct.” Each word carries historical and evaluative weight. None should be attached merely because a version is familiar, technically impressive, slow, fast, associated with a respected name, or available on a popular platform. The listener can still explain a preference. “I prefer the longer phrase in Version A” is a personal response. It is not presented as a documentary verdict.

A recording also cannot supply a raga, tala, composer attribution, or lineage link that it does not state and that the listener cannot otherwise verify. Sound may prompt a question. It does not always answer one. This is why the supporting guide on [how to recognize a raga](https://vidyabhushana.com/blog/how-to-recognize-a-raga.html) should be approached as listening support, not as a licence to assign a final label to every clip. The comparison sheet can preserve a source’s raga label while leaving the identification question open.

The same restraint protects performers. It avoids claiming their intention or accusing them of alteration without evidence. A listener can say that versions differ in a documented passage. The listener should not speculate about why a musician chose a phrase, whether a change was deliberate, or whether a performance represents a wider school. Intent belongs to a source that addresses intent. Where no such source is available, the worksheet should stop at description.

The boundary around recordings becomes clearer when the listener separates **performance fact** from **historical fact**. A performance fact is local: at a marked point in an identified item, this sound is audible, this credit is displayed, or this caption uses this word. A historical fact is broader: a tune was created at a certain time, a text had an earlier form, a practice belongs to a named transmission line, or a composer intended a particular setting. The first category may be well documented by a recording. The second usually needs sources that address history directly.

This distinction prevents a recording from carrying a burden it was never designed to bear. A concert video can be very rich. It may show entrances, phrases, audience context, spoken introductions, visible instrumentation, and a complete rendition. Yet richness of experience is not the same as completeness of provenance. Even a full concert does not automatically tell the listener why a phrase appears, whether another performer inherited it, or whether the text has an earlier documentary witness. The listener should enjoy the richness while keeping the claim local.

Credit information needs the same discipline. A title, a performer name, an event name, and a date might appear in a video description. Those details can be reported as displayed metadata. They should not be expanded into an assertion that the channel controls the rights, preserves the original upload, or supplies every relevant musical detail. If a video has been copied, edited, or uploaded without credits, that uncertainty belongs in the source record. A page that omits a fact has not proved the opposite of that fact.

Short-form material is especially easy to overread. A clipped performance can establish that a sound appears in the clip and that a caption is visible at the time of viewing. It may not include the start of a composition, the conclusion, the source of an arrangement, or the conditions under which the material was edited. A listener should avoid using a clip to make a claim about a whole performance unless the whole performance is actually available. “This clip presents” is a safer verb than “this rendition always does.”

Audio can also be affected by technical mediation. Microphones highlight some voices and instruments more than others. Compression can blur soft consonants and fine movements. A camera edit can change the listener’s sense of timing. A platform can produce a segment from a longer item without making the cut visible. None of these observations requires a technical diagnosis of the file. They simply explain why an uncertain hearing should remain uncertain. A phrase that cannot be clearly isolated is not improved by confidence.

The converse is important too. Do not dismiss recordings because they cannot answer every question. A dated recording can be the most direct evidence for the sound of its own performance. It can anchor a conversation that would otherwise float in memory. It can show a listener where a question begins. It can preserve a present-day instance that later documentation may contextualize. The correct response is not “a recording proves nothing.” The correct response is “a recording proves what it records, subject to its identifiable limits.”

A comparison can therefore be both detailed and restrained. Version A may clearly repeat a phrase at 03:10. Version B may clearly move through the corresponding lyric cue at a different pace. The listener can describe that difference precisely. If neither item names an arrangement, teacher, or edition, the sheet should leave those fields unknown. Specificity in the observed rows and modesty in the interpretive rows are compatible. They are often the best combination.

When a listener wishes to ask a performer, teacher, or archive for clarification, the recording note becomes useful evidence. Share the link, the exact time range, and the precise question. Ask, “Is there a credited source for the text or setting used here?” rather than “Which version is correct?” The former question makes room for documentation. The latter often assumes that the listener’s binary frame is already the right one. Careful source notes improve the quality of the question as much as the quality of the answer.

This principle also applies to silence in a recording. An omitted verse, a shortened repeat, or an instrumental passage can be an audible fact in the recorded instance. It does not by itself show deletion from a textual tradition, a performer’s intention, or a permanent alteration to a song. Treat the audible event as audible. Record what is not present in the item only with that same local qualification. A comparison becomes more reliable when absence is described as absence from a source, not absence from history.

## A three-pass listening method

A three-pass method helps a listener avoid trying to solve every question during the first hearing. The first pass is for orientation. Note the source, title, credit, date, and available context. Identify one or two matching landmarks between the items, such as a recurring word, an instrumental entry, or a clear cadence. Do not fill the comparison columns from memory. The goal is to know what material is actually present.

The second pass is for description. Listen to one short matched passage at a time. Add a text cue only if it is heard or printed clearly enough to record. Add a phrase cue in plain language. Add a tempo observation that is limited to the matched segment. Mark each point with a time range. The notes can be sparse. A short, reproducible note is stronger than a long narrative built from impressions.

The third pass is for context and restraint. Re-read captions, programme notes, edition details, and visible credits. Move contextual claims into the lineage or source column. Ask whether any statement in the worksheet has crossed a boundary. Has a caption label been treated as historical proof? Has a melodic contrast been called a raga change? Has the absence of a credit been turned into a claim that no lineage exists? Correct such moves before sharing the note.

This order also makes the activity more enjoyable. Instead of feeling obliged to identify every detail immediately, the listener can attend to a smaller question. The first pass finds the place. The second pass hears the place. The third pass checks what can responsibly be said about the place. That sequence is appropriate for beginners and experienced listeners alike, because it makes the evidence behind each sentence visible.

Before the first pass, prepare the materials. Open the two sources in separate windows or write their details on a page. Confirm that you are comparing items that genuinely share a usable landmark. A shared title alone may not be enough when titles are abbreviated, spellings vary, or an upload groups more than one item together. If the source does not make the identity of an item clear, write that uncertainty at the top of the sheet. The comparison can continue only as a comparison of labelled media, not as a claim about a single composition.

During the orientation pass, resist the urge to stop every few seconds. Let each item establish its own immediate context. Notice the length of the available extract, whether words are audible, whether there is a visible text display, and whether the item has a caption or programme note. Then choose the most defensible anchor. An exact repeated word is usually helpful when it is clear. If words are not clear, an instrumental entry or a distinct cadence may serve as a temporary anchor. Mark an anchor as temporary when you are not certain it is structurally equivalent in both items.

The description pass works best with short phrases, not whole performances. Listen once for text, once for phrase shape, and once for pace. This avoids writing a broad reaction while trying to catch every detail at once. If the same observation survives each pass, it may be strong enough to keep. If it changes from one hearing to the next, say so. An uncertain observation should not be promoted merely because it is interesting.

Use ordinary language before specialist language. “The voice seems to rise and then linger before the repeated word” is clearer than a technical label that the listener cannot support. If a teacher, score, or source later supplies technical vocabulary, add it with the source. This order protects the listener from using a term as a substitute for evidence. It also lets other readers understand the note even if they do not share the same training.

The context pass should include a small source audit. Ask four questions of each item: Who presents this item? When was it posted or performed, if the source says? What does the source explicitly label? What does the source not explain? The audit often changes the comparison. A phrase that seemed to establish a difference may turn out to be in a cropped clip. A line that seemed to indicate lineage may be a generic biography. A raga label may appear only in a comment rather than in the primary upload. Moving those details into the right column makes the final note stronger.

A listener can also mark the difference between **absence**, **inaudibility**, and **unavailability**. Absence means the feature is not present in the material being compared. Inaudibility means it may be present but cannot be heard clearly in the source. Unavailability means the relevant source, such as a complete recording or an edition, is not accessible in the present comparison. These three descriptions are not interchangeable. Naming them prevents a missing detail from being treated as a musical conclusion.

After the three passes, write a two-sentence summary. The first sentence describes only the strongest comparison result. The second lists the largest remaining boundary. For example: “In the matched refrain, the two present-day recordings have different audible phrase cues and a different relative pace. The materials compared do not explain the transmission, text authority, or reason for these differences.” This short summary is often more useful than a page of speculation because every part has a visible basis.

If the project is shared with others, preserve the raw observation separately from a polished paragraph. The raw note should include links, time stamps, and source wording. The polished paragraph can state the result in readable form. Keeping both prevents later editing from removing the qualifications that made the statement accurate. It also allows another listener to revisit the primary material without having to reverse-engineer the claim.

The method has a final ethical benefit. It trains the listener to present a version as a documented instance rather than an opponent in a debate. This does not deny that differences can matter deeply to musicians and listeners. It means that the public note should identify the difference, give its evidence, and resist claims about people’s intentions or legitimacy that the evidence cannot carry. The result is more respectful, more useful, and easier to correct.

## Build a comparison sheet before interpreting

Interpretation is most useful after the comparison sheet is filled. Before that point, the listener may be hearing a real contrast but attaching the wrong explanation to it. A phrase may seem altered because a different word division changes its shape. A speed difference may make an ornament seem absent. A cropped upload may remove the reference point that would make the two passages comparable. Writing a small ledger first gives these possibilities somewhere to sit.

A good sheet should include a confidence note. “Clear in audio,” “stated by page,” “heard but uncertain,” “not supplied,” and “not comparable” are enough. Avoid a false precision scale if the source does not justify it. The function of confidence language is not to make the listener sound cautious. It tells readers whether they are seeing an audible observation, a visible credit, or an inference that needs more support.

The sheet can also separate a source’s own words from the listener’s language. If an uploader calls an item a particular raga, put that exact label in quotation marks or identify it as an uploader label. If the listener describes a phrase as “descending before the cadence,” mark it as a listening note. This simple distinction prevents borrowed authority from leaking into a personal observation. It also makes later revision easy if the source is corrected or expanded.

Interpretation may then take a modest form. “The present material supports a difference in this phrase and a different stated source context. It does not establish the reason for the difference.” This is a complete conclusion. It is not necessary to fill every blank. In fact, a sheet that displays its gaps can be more useful to a teacher, archivist, or fellow listener than an elegant but unsupported account.

## When a version is described as retuned

“Retuned” is a word that needs careful handling. It can appear in discussion as a present-day description of a rendition. It may point the listener toward a question about musical setting. It does not, on its own, give a complete account of when a setting changed, who set it, what preceded it, or what authority supports the description. The word should remain attached to the source and its context.

The Rasikas.org thread that helped motivate this guide shows why. On 5 September 2026, a participant asked whether someone could decode the swaras of an instrumental interlude in a linked item. [ref-3](https://www.rasikas.org/forums/viewtopic.php?p=386126 "Rasikas.org, Annamayya krithi song swaras, listener request posted 5 September 2026") A reply on 6 September described the item as a song by Annamacharya that “has been retuned in the rAgA kuntalavarALi,” and added that a person well versed in Karnatak music would be needed to write down the swaras. [ref-4](https://www.rasikas.org/forums/viewtopic.php?p=386129 "Rasikas.org, Annamayya krithi song swaras, reply posted 6 September 2026") This is valuable evidence of a dated reader question and of what a forum respondent said. It is not a critical edition, a verified history of the setting, or proof that the forum description settles the matter.

A responsible listener therefore writes, “A forum respondent described this rendition as retuned in Kuntalavarali,” not “the song was historically retuned in Kuntalavarali.” The first statement accurately reports the source. The second adds a historical conclusion that the thread does not document. If another source later provides an edition, a composer note, or a scholarly account, that source can be added to the record. Until then, the comparison retains the distinction.

The same rule applies beyond this thread. A performer, programme, or caption may use the language of setting, tuning, or adaptation. Record the wording, the date, and the speaker when available. Do not fill in an unseen past. A present-day label can be meaningful without being final. The listener’s job is to preserve the label and make its evidentiary boundary clear.

## Raga labels need their own column

A raga label is neither a substitute for a phrase description nor a guarantee that two sources are making the same claim. Labels can be printed in a programme, attached to an upload, stated by a teacher, or inferred by listeners. These routes should be recorded separately. A comparison can include a dedicated **raga label as supplied** column if the label is central to the source context, but it should not silently convert that column into a verdict on musical identity.

This separation is helpful because labels and phrases answer different questions. A label identifies how a source presents an item. A phrase note describes what is audible at a particular location. Two items may carry the same label and still present different phrase cues. Two labels may differ while the listener lacks enough contextual material to explain the difference. The worksheet should preserve both facts without pretending that a short clip resolves the relationship.

The Music Academy abstract itself asks whether pāṭhāntara-s of the same kṛti may be distinctly different when the raga of the kṛti must be considered under different Mēla heads. [ref-1](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract") The value of that statement lies in its question form. It shows that raga treatment belongs among matters that require careful comparison. It does not authorize a listener to assign Mēla heads or raga identities from limited audio.

When labels conflict, note the conflict. Write who supplies each label and where. State whether the compared recordings give any explanation. If they do not, an accurate conclusion may be: “The available sources use different raga labels; this worksheet does not determine why.” That conclusion is more informative than selecting one label because it is familiar. It keeps a document trail open for better sources.

## Editions, notation, and performances are different witnesses

A printed edition, a notation page, and a performance recording may all concern the same title. They are still different witnesses. An edition can show what its editor prints. A notation source can show how it represents a phrase or text. A recording can show what occurred in that recorded performance. None should be silently made to stand for every other form of evidence. The listener should first ask what question each witness can answer.

For text, an edition may be a more direct witness than a compressed live clip, but its own editorial basis should not be assumed. For audible pacing and phrase shape, a recording may be the direct witness, while an edition may not capture every performance nuance. For a stated teacher connection, a programme or performer note may be relevant. Each source has a domain. A responsible comparison uses the sources side by side instead of forcing one to win an argument it cannot answer.

The existing guide to [how to read Carnatic music notation](https://vidyabhushana.com/blog/how-to-read-carnatic-music-notation.html) is useful when a credited score is in hand. This guide does not offer a method for treating any online audio as an authoritative score. A listener may make a provisional contour note from sound. The note should remain a listening note unless a properly credited source provides notation. This is particularly important where ornamentation, text placement, or audio quality makes a pitch-by-pitch account uncertain.

The difference among witnesses also helps explain why apparent disagreement may not be simple contradiction. A score can be concise. A recording can be expansive. A programme can omit details that a performer knows but has not publicly stated. The comparison sheet should record the mismatch in source type. It should not flatten the mismatch into a story of error. Different source types reveal different parts of an event.

## Unknown is a useful result

The most disciplined cell in a comparison sheet is often the one that says **unknown**. It tells the reader that a question was asked and that the available source did not answer it. This is more useful than leaving a blank, because an empty cell can be mistaken for oversight. “Unknown from the available materials” records a deliberate boundary. It says that the listener has not turned silence into information.

Unknown is especially valuable in the lineage field. A source may omit a teacher connection for many reasons. It may be a short upload, a repost, a partial archive entry, or a programme designed for another purpose. The omission does not prove the absence of a lineage. It only means that the source record under comparison does not provide it. The same reasoning applies to text authority, raga labels, tala labels, arrangement credits, and dates.

Unknown can also protect listeners from false historical narratives. Two present-day performances may be different, while the reason for that difference remains undocumented. A listener may suspect that one follows a class tradition and another a concert arrangement. Unless a source establishes that connection, the comparison sheet should not claim it. The note can name the question for future research. It should not answer itself.

There is a positive side to this restraint. When unknowns are visible, better questions emerge. A teacher can point to an authorised edition. A performer can identify a credit. An archivist can suggest a catalogue. A fellow listener can locate a programme note. The worksheet becomes an invitation to improve the record, not a declaration that the record is closed. In a living musical practice, that openness is a strength.

## What the Music Academy lecture frame contributes

The Music Academy’s public programme lists Sangita Kala Acharya Rama Ravi’s lecture, “Pathanthara variations in Muthuswami Dikshitar’s kriti-s,” for 24 December 2025 from 9:05 to 9:50 AM. [ref-2](https://musicacademymadras.in/academic-sessions-16th-december-2025-to-1st-january-2026/ "The Music Academy, Madras, Academic Sessions, 16 December 2025 to 1 January 2026") The associated abstract describes selected Muttusvāmi Dīkṣitar compositions and proposes a practical examination of oral versions in the Dīkṣitar sampradaya. It says the topic is better demonstrated than explained in words and anticipates audio examples where possible. [ref-1](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract")

For this guide, the contribution is a set of comparison questions, not a universal doctrine. The abstract asks whether changes occur in sahitya, melody, or both. It asks about speed variations. It asks whether versions of a kṛti may be distinctly different in relation to raga treatment and whether similar differences appear across other sampradayas. [ref-1](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf "Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract") Those questions justify the five-field listening sheet because they separate kinds of difference rather than treating every contrast as one undifferentiated problem.

The source also supplies an important practical lesson. Audio demonstration matters because a version cannot always be reduced to a prose label. Yet the abstract itself is not a published catalogue of all variants. It does not give a listener permission to generalize about every composition, performer, or lineage. This guide remains within that limit. It uses the lecture framework to improve observation, while refusing to claim that the abstract settles any particular case beyond its stated topic.

Readers who want to use the frame elsewhere should preserve its caution. Start by identifying what exactly is being compared. Separate text from melody, speed from source, and a stated lineage from an inferred one. Then state the evidence limit. The method travels well because it asks modest questions. The answers remain local to the recordings and documents that are actually available.

## What the Rasikas.org thread contributes

The Rasikas.org discussion is not used here as musicological authority. It is used as a dated record of a listener problem. On 5 September 2026, a forum participant linked an item, expressed interest in an instrumental interlude, and asked whether someone could decode the swaras. [ref-3](https://www.rasikas.org/forums/viewtopic.php?p=386126 "Rasikas.org, Annamayya krithi song swaras, listener request posted 5 September 2026") The request is important because it shows how quickly a listener can move from an attractive audible passage to a desire for a settled transcription.

The reply on 6 September took a different position. It described the rendition as retuned in Kuntalavarali and said that writing the swaras requires someone well versed in Karnatak music. [ref-4](https://www.rasikas.org/forums/viewtopic.php?p=386129 "Rasikas.org, Annamayya krithi song swaras, reply posted 6 September 2026") The thread therefore illustrates a useful limit without proving a broader theory. It reminds readers that online audio may be insufficient for an authoritative swara account. It does not prove the history of the item, the correctness of the raga label, or the status of any version beyond what the forum participant stated.

A further response on 7 September said that there were many such items on the forum waiting for someone to decode them. [ref-5](https://www.rasikas.org/forums/viewtopic.php?p=386138 "Rasikas.org, Annamayya krithi song swaras, follow-up posted 7 September 2026") That line can be understood as a participant’s expression of an ongoing community wish. It is not evidence of search volume, representative demand, or an inventory of all relevant repertoire. The distinction matters. Forum activity can document a conversation. It should not be made to carry claims about a whole audience.

Taken together, the posts are a useful editorial prompt. They show why a guide should give listeners a way to make careful notes before asking for certainty. The response is not to discourage questions or close discussion. It is to label the evidence. A listener can ask for a credited edition, seek a teacher’s help, or compare source records. What they should not do is turn a brief online exchange into final proof of historical or theoretical facts.

## How to share notes without overstating them

A useful listening note allows another person to find the same material and understand the size of the claim. Begin with the source: title as displayed, uploader or performer as credited, date if present, link, and time range. Then write one observation in plain language. “At 01:12, Version A repeats the visible lyric cue; Version B moves to the next phrase” is stronger than “Version A has the real text.” The first is checkable. The second asks the reader to accept a conclusion that the note has not earned.

Separate your listening note from material supplied by the source. Use phrases such as “the caption states,” “the programme lists,” “the page labels,” or “the recording audibly contains.” These verbs tell a reader where the statement comes from. If you are unsure, write “I cannot distinguish this passage clearly from the audio” rather than reconstructing it from expectation. A provisional note can still be helpful if it says that it is provisional.

It is also considerate to separate critique from evidence. One can prefer a phrase, wish for a clearer credit, or be curious about a source without declaring a performer wrong. If you ask another listener for help, share the time range and your uncertainty. Ask for a source rather than asking someone to confirm a conclusion. A question framed this way invites correction and preserves respect for the performers whose work is being discussed.

Before publishing a comparison, read every sentence and ask: “Could a reader trace this?” If the answer is no, add a source, reduce the claim, or move the sentence into a question. This small editorial habit matters more than elaborate terminology. It makes space for different versions to be heard with care, and it prevents a listening notebook from becoming an unsupported public verdict.

## Continue listening with supporting guides

This guide is about comparison and documentation. It works best alongside other forms of listening rather than in place of them. [How to listen to Carnatic music](https://vidyabhushana.com/blog/how-to-listen-to-carnatic-music.html) offers a broader entry point for following a performance. [How to recognize a raga](https://vidyabhushana.com/blog/how-to-recognize-a-raga.html) can help a listener hear recurring features, while still leaving a source label distinct from an identification claim. [What gamaka is in Carnatic music](https://vidyabhushana.com/blog/what-is-gamaka-in-carnatic-music.html) gives language for moving notes without requiring a listener to diagnose every phrase.

The guide to [how to read Carnatic music notation](https://vidyabhushana.com/blog/how-to-read-carnatic-music-notation.html) becomes relevant when a credited score is available. It should not be used to force a score out of an unclear clip. The source-led [Carnatic Trinity guide](https://vidyabhushana.com/blog/carnatic-trinity-source-led-guide.html) is a useful reminder that composers, repertories, and sources need defined boundaries. For questions about words and later musical settings in a different repertory context, see [who composed the tunes of Haridasa songs](https://vidyabhushana.com/blog/who-composed-haridasa-song-tunes.html).

The most important next step is simple: choose two clearly identified present-day items, not two memories. Make a small sheet. Record one text cue, one phrase cue, one pace note, any explicit lineage statement, and the source record. Then write what remains unknown. The exercise is successful even if it produces no final verdict. It has trained the listener to hear difference and document it without making claims that the material cannot carry.

For further listening choices and officially presented material, visit the [Living Sabha listening page](https://vidyabhushana.com/listen.html). Return to the source whenever you add a note. A comparison should remain revisable. New editions, programme notes, or performer statements may enlarge the record. The original worksheet remains useful if it clearly shows what it was based on and what it did not attempt to decide.

## A final comparison check

Before sharing a comparison, read it once for hidden verdicts. Replace “the real version,” “the original tune,” or “the corrected text” with a source-bound description unless a suitable edition or archive directly establishes the stronger claim. A clear comparison can be brief: name the two sources, identify one text or phrase cue, note the pace difference if it is audible, and retain an unknown field. The result is not less musical because it is careful. It gives the next listener a repeatable path back to the materials that prompted the question.

## Frequently Asked Questions

### Is one pāṭhāntara automatically more correct than another?

No. This guide does not assign a correct or incorrect label to a version. It asks the listener to document what differs in a specified source. A conclusion about textual authority, transmission, or historical priority requires evidence appropriate to that claim. Familiarity, reputation, or a convincing performance does not by itself supply that evidence.

### Can I use a YouTube or social-media clip to identify the original tune?

A clip can document what is heard and what its visible caption states. It cannot by itself establish an original tune, a composer’s intention, or a full history of transmission. Record the clip’s date, credit, and time range. If the question matters, look for a credited edition, programme note, archive record, or an informed source that addresses the claim directly.

### What should I do when two sources use different raga labels?

Record each label with the source that supplies it. Do not remove the disagreement by selecting the more familiar label. Compare the audible passage only as far as the material allows, and state if the reason for the differing labels is not supplied. A source label and a final musical conclusion are different things.

### Does a difference in speed prove a different lineage?

No. Different pace in matched passages is a listening observation. It does not, by itself, establish a lineage, a historical change, or an error. Record the compared location and describe the pace carefully. Add a lineage statement only if a reliable source explicitly provides one.

### A forum member called a rendition retuned. Can I repeat that as fact?

Repeat it as a report of what the forum member said, with the date and link. Do not expand it into a settled historical assertion unless another suitable source supports that assertion. The Rasikas.org thread discussed in this guide is evidence of a conversation and a stated view, not a critical edition. [ref-4](https://www.rasikas.org/forums/viewtopic.php?p=386129 "Rasikas.org, Annamayya krithi song swaras, reply posted 6 September 2026")

### Why keep an “unknown” field instead of leaving it blank?

A visible unknown tells the next reader that the question was considered but not answered by the material available. It prevents silence from becoming an implied claim. It also shows where a teacher, performer note, edition, or archive might improve the comparison later.

### Can I transcribe the swaras I hear?

You can make a provisional listening note, such as a contour description and a time range. This guide does not treat that note as authoritative swara notation. The forum response cited here similarly stressed the need for strong Carnatic expertise when writing down swaras from the item under discussion. [ref-4](https://www.rasikas.org/forums/viewtopic.php?p=386129 "Rasikas.org, Annamayya krithi song swaras, reply posted 6 September 2026") Use a credited notation or qualified guidance for a formal transcription.

## References

1.  [Rama Ravi, Pāṭhāntara variations in Muttusvāmi Dīkṣitar’s kṛti-s, Music Academy lecture abstract](https://musicacademymadras.in/wp-content/uploads/2025/12/24.12.2025-9.05-9.50am-Patanthara-variations-in-Muthuswami-Dikshitar-kriti-s.docx.pdf)
2.  [The Music Academy, Madras, Academic Sessions, 16 December 2025 to 1 January 2026](https://musicacademymadras.in/academic-sessions-16th-december-2025-to-1st-january-2026/)
3.  [Rasikas.org, Annamayya krithi song swaras, listener request posted 5 September 2026](https://www.rasikas.org/forums/viewtopic.php?p=386126)
4.  [Rasikas.org, Annamayya krithi song swaras, reply posted 6 September 2026](https://www.rasikas.org/forums/viewtopic.php?p=386129)
5.  [Rasikas.org, Annamayya krithi song swaras, follow-up posted 7 September 2026](https://www.rasikas.org/forums/viewtopic.php?p=386138)