Objections / Issues On The Integrator Quizzes

Use this discussion to ask questions / raise objections on the 2 integrator quizzes.

From David:

"Tech Quiz

Question 1 cant really be answered unless the specifications of the proposed cameras are known. With the lack of information there are two answers that could be correct.

Question 7 What is this question trying to identify? I would expect similar results from similar settings but without knowing more details this question is meaningless. The lack of information also leaves more than one answer that could be considered correct.

Question 12 The correct answer given for this, in my opinion, is a perfect world scenario. Other factors not specified could play a part in this answer."

David, for question 1, please explain what two answers you think would be correct (e.g., thermal, super low light, integrated IR, etc.). Thermal is not going to give 'the most details possible' so that is out. Super low light cameras at a measured 0.1 lux in a scene is not going to give many details at all. By far, integrated IR will give 'the most details'.

Question 7 is asking about how to measure / compare compression levels. You would be wrong to expect similar results from similar settings. We've tested this extensively. See: IP Camera Manufacturer Compression Comparison

Please expand on your take on Question 12. What scenarios? What answers do you think are correct?

Short on time at the moment but not knowing the specifications of either the super low light or IR camera, either could provide a better image over the other depending. This is what I had the problem with. Low end IR cams as I know you've worked with can create issues such as hot spots etc causing the image to wash out or poor illumination paterns. Just the fact that the choices are generalized to IR camera or super low light without defining those invalidates it in my opinion.

The key detail is 0.1 lux. We've tested a range of new high end super low light cameras and they all struggle greatly at 0.1 lux (real measurement, not a make believe lux rating).

Keeping in hot spots or wash out is smart. That said, at 0.1 lux, overwhelmingly integrated IR cameras will deliver more details.

I will concur with the quality of the cameras being equal and high end, IR would win. Maybe adding a similar statement about equal quality to qualify the question better. I was also taking into consideration that the super low light would be using a sens-up feature or similar. Not being in the field anymore and no access to a lux meter I also was basing judgement on non scientific testing and my real world experience.

The problem I had with 7 is there isnt enough information about what the scales relate to. Basing off your reply and link to the comparison I understand where you were going but with only the information presented in question and pictures of two scales chosing the answer the scales are the same with different terminology is accurate in my opinion.

I will remove my disagreement with 12. You did ask what is best and the answer marked as correct is the best if you have access to the interior structure. There are other factors that would be considered when chosing a best solution for the application such as weight and environment in my opinion. I would think that adding information about exactly what you are hanging on this wall would help clarify. ie small, medium, larg speed dome, bullet, box cam with wall mount and housing etc.

I appreciate the discussion.

"with only the information presented in question and pictures of two scales choosing the answer the scales are the same with different terminology is accurate in my opinion."

David, it's important that surveillance professionals recognize that compression scales are not commensurable / comparable across manufacturers. It's a common misperception but testing shows that significant variations across manufacturers exist in what quantization / compression levels manufacturers default to and how they vary their scales.

The practical impact is on bandwidth / storage needs and choosing across different cameras. Because of these variations and scale differences, it's easy to be misled that one camera saves a lot on bandwidth / storage, simply because the scales are set differently.

I understand your points. I would think it appropriate, so there is no confusion, the question reflects that the tables come from two different manufacturers. As it is we are left to assume that. Regardless of how safe an assumption that is we all know what assuming gets us.

For example, the two tables could have come from the same manufacturer but different versions of firmware or even software and the table was revised.

David, good feedback. I've changed the question wording to "Which of the following statements is most accurate when comparing two different camera manufacturer compression scales?" This should make that aspect explicit. Thanks!

question number 8:

we need to know what power consumption per camera to calculate this. 802.3af offers 15.4w per channel but if your 12 cameras only draw 5W per camera (for example) then the total power budget of the 802.3af switch just needs to be 60W.

The question is: "What should your 802.3af PoE switch's power budget be in order to ensure 12 IP cameras will work without issue?"

You don't know / aren't given the power draw in this scenario so, in order to 'ensure' it will work without issue, you need to supply the max power supported by 802.3af.

I agree with 'A'. Even with my limited IP camera experience, I've seen many cameras that require only ~5W. Then again, I've seen IP cameras with power zoom/focus that require up to 15W (and if used on longer runs or with extenders, over 15W, requiring Class 3 PoE) so 70W may be more than sufficient or 120W may not be. Ambiguity again...

Axis 'P' series, for instance, requires something like 14.6W, if I remember correctly, so 12 of them would require 175W - 180W would not be sufficient to allow any overhead. And you didn't specify fixed vs. PTZ, which is an entirely different ball game.

"How many SD cameras can a 5 megapixel camera replace?"

I think the use of 'can' to be problematic here, since its not clear what 'can' means here, i.e., 'in theory' or 'at most' or 'ever'. Reversing the question also helps show why, e.g.

True or False: A 5 MP can replace 16 SD cameras. True, it could if everything was perfect.

But just changing 'can' to 'will' fixes the problem. Sorry for the double post, but I feared the first one wasn't clear and it didn't actually reference the original question.

That question has 4 answers listed: 4, 16, 20 and 'not enough information'

Can anyone seriously argue that any of the answers above besides 'not enough information' is correct here? If so, please explain your logic for choosing 4, 16 or 20 over 'not enough information'.

The whole point of this question is to demonstrate that the person is sophisticated enough to understand that how many cameras can be replaced depends on more than just the resolution / pixel count.

Can anyone seriously argue that any of the answers above besides 'not enough information' is correct here?

I am unable to come up with more than three ways:

1) The statement 'not enough information' implies that IF we hand all the information, then we would decide sometimes that, yes, 4sd cameras are able to be replaced by a single 5mp camera, in certain scenarios. But if that is true, even in one scenario, then it makes the statement: "4sd cameras can be replaced by a 5mp camera" true as well...

Always can? No. Sometimes can. Yes. The word 'can' means 'sometimes can' by default.

2) Parallel construction: How many feet away can John Honovich sink a put from?

4ft, 16ft, 20ft, Not enough information. Let's say we played a round, I saw you sink a 16 ft put one hole. And you didn't ever sink a 20ft put. From 4ft away John Honovich can sink a put. This is true, Yes /No?

3) FWIW: Consider that when answering the question I was fully aware that this question was trying to show that there's more than just adding up the pixels to consider. I looked at it for 45 seconds, and yet I was compelled to choose what I knew wasn't what you thought right answer was. Just because of 'can'.

But if you just change 'can' to 'always can' than its clearer for sure, right? So what's the big deal?

Rukmini -

Your arguments underscore the difference between *knowing* information and *understanding* it.

Yes, you could twist things a lot of different ways to argue valid answers for that question.

Like any test, the questions and answers are generally going to be in context of the person or organization administering the test. If you've read IPVM for a while and you understand some of what I would consider the baseline arguments or counter-arguments presented here, I think the correct answer to the test is fairly obvious.

I didn't feel that question and answer selecion had any ambiguity at all, as long as you had "done your homework" going in to it. Of course, if you want to argue for arguments sake, there was tons of material available.

B, you seem like a logical undisclosed guy, so I will, as Marty wisely says, drop all the minor charges and focus on one point:

Direct question: Do you agree that after changing 'can' to 'will' the correct answer remains 'not enough information'?

Is anyone seriously going to argue that 'will' is less clear than 'can'?

'Can' is known problematic word in propositional logic, 'could' or 'will' are the most common replacements.

Again, I'm not saying the question is bad, I'm just suggesting how to make it better. Why the entrenched response? We're all friends here, right?

Honestly, I don't think your wording substitution is any better, it just trades one set of ambiguities for another. I was going to write out a longer answer, but I've already spent too much time on this.

Bottom line, I don't think your suggestion makes the question better in the grand scheme of things, it only makes it different.

it just trades one set of ambiguities for another...

John can play in the golf tournament. Maybe he will and maybe he will not.

John will play in the golf tournament. [Fill in ambiguities that don't apply to can]

Apologies B, I was blinded by logic. You gave me the hint and I just ignored it:

Like any test, the questions and answers are generally going to be in context of the person or organization administering the test. If you've read IPVM for a while and you understand some of what I would consider the baseline arguments or counter-arguments presented here, I think the correct answer to the test is fairly obvious.

As I'm sure you have known all along, the word 'can' IS fundemental to the question. Why? Because it is the exact language used by the vendors who would try and swindle us! Saying 16sd cameras 'can' be replaced, would be tantamount to saying Avigilon's claim of 'you can replace 95 cameras with single 29MP one' might be true! Never!


You need a hobby.

Besides this.


#6 on the tech quiz asks which component "most likely impacts bandwidth handling capability of a server?"

GPU is not wrong - as it most likely impacts the memory bandwidth handling capability of a server.

ok, maybe in my scenario the end user is using Client on a server to view video because that's all they have.

weakens my point some but does not nullify it. :)

So your argument for a *server* is that the worlds best GPU coupled with a Pentium II would outperform an off-the-shelf i3 machine in bandwidth handling as it typically relates to a server for IP video?

my argument is that GPU is not wrong on #6 in the tech quiz

So, of those componets, if you had to pick a single one you think the GPU is the *most likely* to impact bandwidth? Not semi-likely, not a contributing factor, but the MOST likely to impact bandwidth handling?

I'm surprised that GPU was even listed. I would have liked to see "SATA 1 Drive" and "10Mbps NIC" as options :)

The question did not ask for most likely.

Unless it's been edited recently, this is the exact text of the question:

"Which of the following PC components is most likely to impact bandwidth handling capability of a server?"

Either way, your answer sucks ;)

Wait a second, B. ACC 5 has implemented server side CUDA, GPU based H.264 decoding, used independently of whether there is 'local monitoring' or not. According to this VMS/CUDA (not related directly to ACC 5) research, with sixteen streams, the CPU load is at 13% with GPU enabled vs 100% with no GPU.

So the *correct* answer should be fairly obvious, 'not enough information', right?

No. So assume you don't have a GPU. What component do you need to scale up? CPU. Which makes it still the correct answer.

Long story short, we're not changing this. Let's move on.

It WAS edited recently... my words were copied and pasted from the original.

the addition of the word 'to' (and the loss of an 's' on impact) - which did not appear in the original - now makes GPU wrong.

Reread my words. I never said GPU was the 'best' answer (because the original question didn't ask that) - just that GPU wasn't wrong.

Now it is, and I (and I hope, you) will commence living happily ever after.

CPU is clearly the most common / fundamental component impacting bandwidth consumption. This is not only shown by our testing but is the wide consensus of VMS / recorder engineers.

Of the 4 choices, here is how test takers responded:

  • 69% CPU
  • 20% RAM
  • 11% GPU
  • 0% Power Supply Wattage

The larger issue is people saying RAM, because that's clearly a misconception. GPU is at least related / impacts performance but the question did ask for 'server' and 'bandwidth' so picking GPU over CPU in this context is hard to justify. Maybe if the question was client side viewing....

On two of the ones I missed, I had first selected one answer and then changed my mind and selected another. The test results picked up my first answer which was incorrect but my final answer had been the correct ones. Not sure how/why that would happen but might be worth looking into to make sure final answers are recorded? Specifically this was on Sales quiz #2 and #10. In any case, its clear I need more training! :)

Minor possible correction to Surveillance Integrator Sales Quiz

The first question is

1 1. You propose a camera with a .01 lux rating and your rival proposes a camera with a .000005 lux rating. The end user raises a concern about low light performance. Of the following options, which is most accurate?

the last answer choice is

A third camera with a lux rating better than .00005 will ensure that you have better low light performance than you competitor.

Which is not correct in any case, but may give one the mistaken idea that had it said .000005 (like in the question) instead of .00005, then it would have been correct.