[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

IBC96 Telecine observations




While I would not normally express my opinions on the Internet I feel that
misleading information winging it's way across this media deserves some
comment to set the record straight and may also explain why we at Cintel
have "ruffled our feathers" more than usual.


I was fortunate enough to visit  Tape House New York in August this year
just before IBC 1996. 

During my visit John Dowdell showed me two tapes both of a 16mm IP
documentary.

One tape was an Ursa Gold transfer the other a Spirit transfer.

John Dowdell quite understandably enthused about the Spirit transfer and
indicated that the Ursa Gold transfer was the best that Tape House had
managed to achieve, although he himself had not done the transfer.  Present
at this showing was the  colorist who, I understood, had done the Ursa Gold
transfer.

I suggested that I thought Ursa Gold could do a better job of the 16mm IP
than I was being shown. John Dowdell and the colorist invited me to look at
this 16mm IP on the same Ursa Gold as originally used.

We spent about 10 minutes looking at various shots, and John Dowdell
identified a shot that looked incorrect on the Ursa Gold transfer but good
on the Spirit transfer. The shot included Red roofed houses, sky and trees
with foliage.

The Ursa Gold version showed brown rather than green foliage and somewhat
undersaturated red roofs. Additionally the foliage detail appeared sharper
on Spirit.   
We put the shot back on Ursa Gold and using only Ursa Gold color correction
proceeded to match the Spirit transfer.

Without using any Ursa Gold IP masking we balanced PECs and with admittedly
a considerable gain offset and some differential gamma managed to get a
color balance  that gave green foliage while holding Red roofs and correct
skies. At this stage color saturation was lower than the Spirit transfer,
by increasing the Ursa Gold saturation control by about 30% a very
acceptable colorimetric match was achieved.

The foliage resolution was still somewhat lower on Ursa Gold and was then
matched to the Spirit resolution by adding under 10% aperture correction.

Noise on transfers was not noticed to be that different.

Steadiness was similar on both transfers, and being a rotary IP any
unsteadiness noticed in either transfer could be attributed to the printing
stage.


Both John Dowdell and his colorist agreed that a near, if not exact match
between the 2 telecines had been achieved, admittedly with some more
adjustment required on Ursa Gold but nevertheless without any downstream
color correctors. 

(We are also presently checking whether I.P. is correctly "woken up" at
Tape House by their telecine controllers as I personally found the results
a bit odd compared to other Ursa Golds; despite a good match being
achieved.) 

I was therefore saddened to find at IBC96 that John Dowdell was showing a
split screen demo of the original Ursa Gold transfer against Spirit and
implying that that was the best one could expect from Ursa Gold, especially
when taking into consideration the above.

I believe that Ursa Gold can achieve similar quality to Spirit for IP
transfers, despite the fact that the Spirit front end was originally
designed to cater best for IP while Ursa Gold is designed to encapture any
material from negative through IP to the darkest projection print.

                       
------------------------------------------------------

It is a shame that technical comments by Cintel became confused at IBC
1996, especially those relating to other material that our friends Philips
were demonstrating.

So here I would like to put the record straight.

References made by Cintel to Spirit steadiness became confused by certain
staff members. 

We do not contest the excellent steadiness of 16mm on either Spirit or Ursa
Gold.

We did observe that the 35mm Spirit steadiness demo was inconsistent as
single layer steadiness grids showed minor weave whereas a composite of
these layers appeared steadier than the individual shots. I would dearly
like to understand how this was possible without further downstream
processing at the compositing stage.

We have also observed at IBC and elsewhere that the 525 & 625 35mm Spirit
transfers, while obviously showing less aliasing than Ursa Gold, seem to
loose some fine image detail compared to Ursa Gold.

While the film's grain structure is less prominent on Spirit transfers
there seems to be some evidence, a strange image lag, that this may be due
to built in noise reduction which may also play a part in this slight
reduction in observed resolution. 



Giving my opinion of Spirit at 2K compared to the same film on our Klone at
2K I have to say that the demo on the Philips booth looked soft and lacking
in "punch", which the Philips demonstrator said was due to it being an
internegative, which it was obviously not, (Kodak confirmed that the
material was original negative), while the Thunder on the Kodak booth using
again the same film looked sharp but noisy, rather than film grainy, (noise
at these resolutions looks neutral while film  grain is multicolored)  the
Thunder looked almost over apertured with some weird coring artifacts,
explained honestly by Kodak as a fault in the Thunder, yes we all have
embarrassing moments. [I remember the first URSA demo in the USA, it caught
fire, ahh those were the days]


I have tried to be objective here and only state my observations. We are
not perfect, many things can and will improve with our telecine technology,
no doubt so will  Philips technology; however I guess I have an unfair
advantage in knowing what's coming next from Cintel, and I while I cannot
offer previews of the future I can say that what I saw at IBC 1996 from all
manufacturers of telecines and scanners including ourselves does not match
what the future holds.

You ain't seen nothing yet !!

P R Swinson

1 October 1996