DCP2K4 w/dolphin to Barco Series 4 4K

[Leo Enticknap]

Setting aside the math related to throws and screen sizes, lumens measures what comes out of a projector. Foot-lamberts measures what reaches your eye (or the eye of the color meter that your service tech will show up with), after the lens, port glass, ambient light in the auditorium, tivoli lights, exit boxes, customers' smartphones, and, crucially, the reflectivity of the screen surface have all taken their cut. That's why in the cinema world, the lumen figure is not all that important: the ft-l figure is what matters, because DCI says that we have to hit that magic 14.

What brought all this home to me was an install in a mixed use venue just after covid. They mainly did live shows but wanted to up their game in arthouse cinema. I put an SP2K-11C in for them. The numbers we ran ahead of time said that it should have given them 14 with 2-3 of headroom. It didn't: with the lasers maxed out it was barely hitting 11 in flat (common width screen), and the image looked dull (lacking contrast and general "pop"). After some thinking about it, I realized that the "screen" was a cyclorama, not an actual cinema screen surface. The budget wasn't there for a Stewart Snomatte or any obvious solution of that sort. This was what we came up with:

image.png

Worked like a charm. The projector is high enough above the seats and quiet enough that fan noise isn't an issue, and we hit 14 on around 80% laser power. Regular window glass sure does drink a lot of light.

[Steve Guttag]

It eats it in two way. It isn't as transmissive as people think (And that is when new/clean)...it greens up the image (iron content in the glass) so you lose more light when you color correct. Even good water clear glass is going to suck 8%-10% of the light. AR coating is key too. Any light you see reflected on the back wall or floor (or ceiling, I suppose) is light not reaching the screen. FWIW...even presuming proper glass, I presume 10% loss on the glass and another 10-15% on color correction. One is making a very long gamble on laser light. Many specifications call out laser life until it is at 50%...well, if you don't plan on your decay to that level, you are going to get substantially less life. The harder you run the lasers (the hotter they get), the less life you get.

[Frank Cox]

That leaves me wondering why projector installations don't have lenses or lens covers that would somehow attach to a hole in the port glass without leaving a gap between the lens casing and the glass.

Then you would get the best of both worlds: the light would be unobstructed by the glass and the mechanical noises would still be confined to the booth.

[Ryan Gallagher]

Worked like a charm. The projector is high enough above the seats and quiet enough that fan noise isn't an issue, and we hit 14 on around 80% laser power. Regular window glass sure does drink a lot of light.

Haha, little mouse hole. We may have to consider such options. But if I can keep the 2K off, the booth is probably quiet enough we could just remove that port glass. (it's normally used as a spotlight position).

Steve, I got curious and bumped the barco calculator up to maximum porthole and other losses. (25% and 50%)... only in the final % adjustment did it bump us up to the 15C recommendation. So we should have plenty if they arrive with a 15C, even with crappy porthole and extra color correcting involved I hope. I had lifetime set at minimum because this is a rental.

New Lossy Calculation

I think there was at one point a 3D title on our festival schedule that went away, and that might explain the 15C over the 12C. At more modest port and other losses it recommends the 15C for heavy 3D use. As a permanent install with a more typical lifetime value, it recommends the 25C for lots of 3D. Maxed lifetime (70k hours) with varying % of 3D shows range from the 20C to the 35B recommendation. This particular projector may also have a history and be mid-lifetime somewhere.

I was aware of running below full laser power being a desire for permanent installs, and knew 3D was a factor too, but was not aware how much massive of a swing it could produce in the projector solutions (and costs).

[Ryan Gallagher]

That leaves me wondering why projector installations don't have lenses or lens covers that would somehow attach to a hole in the port glass without leaving a gap between the lens casing and the glass.

Then you would get the best of both worlds: the light would be unobstructed by the glass and the mechanical noises would still be confined to the booth.

Fun thought, but that kinda relies on a perpendicular shot... unless you have some clever attachment for the variable angles. It's also pretty hard to swap a lens that is attached to the port glass! Ours would be a nightmare even without that, two periscope mirrors to remove and reset. Another bonus of keeping it all in the booth is the booth air can be filtered/managed better than the house air. We have so much atmospherics in our non-film shows that there would need to be a cap that fit on the house side of such a solution too, otherwise you'd be contact cleaning a very expensive lens weekly.

But if they are really all 10-20% lossy, seems like maybe there is room for such an approach. In some cases might save you a whole tier in projector spec.

[Leo Enticknap]

That leaves me wondering why projector installations don't have lenses or lens covers that would somehow attach to a hole in the port glass without leaving a gap between the lens casing and the glass.

As an alternative to optical port glasses that cost hundreds of bucks for an 18 square inch pane, it should be perfectly possible to come up with a snood that fits into an existing port frame and attaches to the projector's lens. Slightly different designs would be needed for the various lens diameters. They would likely need to be telescoping, and hinged on the X and Y axes, to give you a bit of wriggle room for the positioning of the projector pedestal. But I can think of no insurmountable barrier to designing one, and that is noise resistant enough that it prevents any significant unwanted noise from reaching the auditorium.

[Steve Guttag]

What would work is to have a set of bellows that, after installation/servicing could be pulled up to the projector and attached to, in effect put the lens area of the projector in the theatre space. Then, when you need to service/clean disconnect the bellows.

[Ryan Gallagher]

Brief update after install. No snood or glass cutter required. Ed from Barco showed up with a DP4K-25C.... tons of output headroom. We are at 14fL through our crappy glass with just 43% laser power.

Install was not without challenges. Pedestal had to be mostly disassembled to fit through the booth door. Projector barely made it in upright orientation but didn't have to flip it 90.

The screen geometry is the one real compromise on the install due to it playing 2nd fiddle to 3 existing projectors in the booth... I'm gonna have some "fun" making barco picture masking macros to account for the festival's typical slew of unusual aspect ratios. No major complaints during the two tech checks today... but if anyone does any historical photos or footage set within the frame it is going to be painfully obvious we are not dealing with ideals here. Far from ideal, but will probably get us through the festival with only a couple serious complaints. We have a better shaped 2K image to offer them if so.

Showed up with an ICMP-X and of course festival is doing a lot of late in the game key acquisition, but they at least had a few days heads up on the target serial.

Still getting keys for the Doremi too so doing double work this fest making sure it is a viable backup.

[Ryan Gallagher]

LOL right before opening show we realized the festival walkin slides were the least acceptable in the offset geometry. Had to fire up the 2K just to run slides. Been too busy to even look at them and they always deliver those in the 9th hour because of last minute schedule changes.

[Leo Enticknap]

If you have keystones, cropping or generally wacked up projection geometry caused by a big down angle and/or the projector being horizontally way off the centerline of the screen, slides are what is most likely to make that visible to the untrained eye, especially if they contain text close to the edge of the frame. The other big danger is subtitles, for the same reason.

[Ryan Gallagher]

If you have keystones, cropping or generally wacked up projection geometry caused by a big down angle and/or the projector being horizontally way off the centerline of the screen, slides are what is most likely to make that visible to the untrained eye, especially if they contain text close to the edge of the frame. The other big danger is subtitles, for the same reason.

Oh it's all kinds of bad with certain content. Don't look at the crawl center-line too closely, don't watch the text emerge from behind the bottom masking. Look away when there is an AR transition within the film itself, no amount of cropping can fix that unless this was my permanent projector and I had way more tech time to put internal macros within the feature. Yuck.

I don't think it was worth it without geometry features. Barco is obviously keen to be in the running for our Series 2 replacement. But this was not the kind of demonstration they would have hoped for... they should have tried harder convicing us to kick the 2K out of it's near center placement. But they also wanted to load out the sponsor projector the same day as the festival awards show that still needed a projector, kinda killed that ideal placement idea. If we had it till the true end of the festival I would have been more than happy to spend a day restoring and checking the 2K. Alas.

[Ryan Gallagher]

How happy were they with the ICMP-X's ingest speeds? The new ICMP-XS is to be speedy but the ICMP-X...not so much.

Our FTP/network ingest speeds seem fine. We used a separate GBit patch from the ICMP-X to our booth switch (which has adequate bandwidth). It can ingest over just the management network interface but Barco advised against that because you'll end up sharing the bandwidth on the internal switch.

Our ICMP-X USB and CRU speeds however are minimal improvement over the Doremi. The ICMP-X has two USB 2.0 ports and two USB 3.0 ports. For whatever reason I have not had a chance to determine, none of our USB devices recognize on the 3.0 ports, only the 2.0 ports. So bottleneck is the same there as it was with the Doremi, even though the ICMP-X has SSDs for it's storage array.

Maybe that's solvable, if anyone knows the tip i'm all ears, but unlikely to have too many more CRUs this festival. My sled supports E-Sata too, I'll have look in the dark corner again to see if that port exists.

Another glitch is some but not all of our SimpleDCP server delivered CPLs keep showing up as if they are not ingested every time you connect to that source. You can re-ingest them and it "completes" almost instantly, but never marks them as ingested next reload. Makes keeping an eye out for the fresh deliveries a bit annoying, cause lots of noise in the list.

[Leo Enticknap]

With the ICMP (all variants), you have to uncheck the "quick scan" option on its web UI's ingest page, or else it will fail to see some CPLs entirely, and bork out while ingesting others. This has been the case for as long as I've been using and installing them (since 2015), with all three hardware variants, and all the software/firmware versions. As for the USB3 ports, they will see flash sticks and externally powered USB HDDs without any problem, but don't provide enough current to run many if not most USB-powered HDDs. If you regularly have to ingest from USB-powered HDDs, an externally powered USB3 hub is essential.

The network ingest via the management LAN issue is because if a high bandwidth stream goes in through the CCB's NIC, it then has to go across the backplane into the ICMP, and this can overwhelm the backplane's bandwidth capacity and freeze up the entire projector. Obviously if you're going straight into one of the NICs on the ICMP, that isn't an issue.

[Ryan Gallagher]

With the ICMP (all variants), you have to uncheck the "quick scan" option on its web UI's ingest page, or else it will fail to see some CPLs entirely, and bork out while ingesting others. This has been the case for as long as I've been using and installing them (since 2015), with all three hardware variants, and all the software/firmware versions. As for the USB3 ports, they will see flash sticks and externally powered USB HDDs without any problem, but don't provide enough current to run many if not most USB-powered HDDs. If you regularly have to ingest from USB-powered HDDs, an externally powered USB3 hub is essential.

The network ingest via the management LAN issue is because if a high bandwidth stream goes in through the CCB's NIC, it then has to go across the backplane into the ICMP, and this can overwhelm the backplane's bandwidth capacity and freeze up the entire projector. Obviously if you're going straight into one of the NICs on the ICMP, that isn't an issue.

Thanks Leo.

Things we tried in the USB 3.0 port with no love including: a regular flash stick, my Crucial X10 Pro SSD, both Dolby and Barco tech flash drives, and a wall powered CRU enclosure.

I'll try again tomorrow when I am back because I did notice sometimes it takes longer than you would think to recognize a new device. The quick-scan tip is good too, maybe that will unclutter our ingest interface or improve the 3.0 visibility.

[Leo Enticknap]

That's interesting. Last month I installed an SP4K-15C with an ICMP-X in it: a week later, the end user reported the same symptom. USB3 sticks would sometimes fail to be recognized, and a reboot of the ICMP-X was needed to fix that; but the fault never presented when using the CRU reader (externally powered, obviously) via USB3. I therefore put it down to the known issue of the relatively weak USB power supply in the ICMP (all variants), but am now wondering if there could be a software/firmware bug at work.