Details

Published: 17 May 2026

Last Updated: 17 May 2026

A typical outcome of a discussion on RMIM Whatsapp group. 

Sandeep Dougal reports:

When did Pankaj Mullick record his versions of the three songs KL Saigal sang in Meri Bahen?

A question came up recently about whether Pankaj Mullick’s “version” recordings of the three Meri Bahen songs may actually have been recorded later, possibly around 1948 rather than contemporaneously with the 1944 film itself.

To try to understand this better, I put together whatever Columbia VE-series catalogue and CEI matrix-number data I could readily locate from surviving label images, Discogs entries, the HMV India listings, and M V Surender's work on this website

Columbia assigned CEI matrix numbers sequentially as recordings were made. By mapping known CEI numbers against releases with established dates, one can broadly bracket when particular recordings were made.

The relevant sequence is summarised in the accompanying table.

The Pankaj Mullick Meri Bahen versions were issued on VE 2526 and VE 2527 with matrix numbers CEI 23484, 23485, 23486 and 23487.

These sit cleanly between:

Wapas (1943): CEI 22512 / 22514

and

Hamrahi (1945): CEI 23929 / 23931

The nearest confirmed 1948 release I could readily locate in the same VE sequence was Anirban (VE 2553), carrying matrix numbers CEI 27408 / 27409, nearly 4000 numbers higher.

For a 1948 recording date to hold, the Columbia CEI numbering system would therefore have had to move backwards by nearly 4000 numbers, which would be highly unusual in a sequential matrix-numbering system.

That said, matrix and catalogue sequencing in the recording industry was never mechanically linear in every detail. Recordings from different sessions, and sometimes even different years, could later be paired together on the same disc depending on commercial considerations, festivals, anniversaries, market demand and so on. Such local inversions and mixed pairings occur routinely across recording-company catalogues, including within the VE sequence itself.

What nevertheless remains striking here is the broader chronological clustering. The Pankaj Mullick Meri Bahen versions belong firmly within the 23XXX recording environment of the mid-1940s rather than anywhere near the 27XXX range represented by the confirmed 1948 issue Anirban.

The matrix evidence therefore points quite strongly toward contemporaneous wartime recording, i.e. around 1944 alongside the film itself.

'Could Mullick recordings precede Saigal's versions used in the film?'

Only surviving Columbia recording logbooks could establish exact session dates with absolute certainty. That would become particularly relevant if one wished to determine whether the Pankaj Mullick recordings may actually have preceded the KL Saigal versions, as some accounts suggest that Mullick had originally been envisaged for the role later played by Saigal. 

Even that question, however, would require careful reconstruction across different company and matrix systems, since the Saigal and Pankaj Mullick recordings belong to different recording series.

In the absence of such surviving recording logbooks, however, matrix-sequence analysis of this sort is the standard method long used by discographers including Michael Kinnear.

What the present VE / CEI evidence does seem to establish fairly strongly is something narrower but still important: that the Pankaj Mullick versions themselves belong to the same broad wartime recording environment as the 1944 film, rather than representing substantially later postwar recreations.

'Could the Mullick recordings have been released later?'

My primary interest when I first came across the Pankaj Mullick versions was in establishing when they were recorded, not when the discs may have been commercially released. The separate question of release chronology arose only because someone wondered whether the records may perhaps have appeared later, after Saigal’s death in January 1947, possibly around 1948. However, the VE catalogue sequence itself seems to point otherwise:

VE 2526 and VE 2527 sit naturally within surrounding 1943–45 issues, while the confirmed 1948 issue Anirban already appears later in the sequence as VE 2553. Unless Columbia was assigning catalogue numbers radically out of order, the catalogue evidence too points toward a mid-1940s issue rather than a post-1947 release.

A broader sweep through additional VE 25XX issues would likely make the release chronology even clearer.

Sources:
– This website
– Discogs
– piezoelektric.org HMV India VE-series listings

The KL Saigal and Pankaj Mullick versions are available on Youtube.

Now You Know! 

New DS-680